however, do not affect the
conclusion that there is for each planet a certain specific velocity
appropriate to that body, and depending solely upon its size and mass,
with which we should have to discharge a projectile, in order to prevent
the attraction of that body from pulling the projectile back again.

It is a simple matter of calculation to determine this "critical
velocity" for any celestial body. The greater the body the greater in
general must be the initial speed which will enable the projectile to
forsake for ever the globe from which it has been discharged. As we
have already indicated, this speed is about seven miles per second on
the earth. It would be three on the planet Mercury, three and a half on
Mars, twenty-two on Saturn, and thirty-seven on Jupiter; while for a
missile to depart from the sun without prospect of return, it must leave
the brilliant surface at a speed not less than 391 miles per second.

Supposing that a quantity of free hydrogen was present in our
atmosphere, its molecules would move with an average velocity of about
one mile per second. It would occasionally happen by a combination of
circumstances that a molecule would attain a speed which exceeded seven
miles a second. If this happened on the confines of the atmosphere where
it escaped collision with other molecules, the latter object would fly
off into space, and would not be recaptured by the earth. By incessant
repetitions of this process, in the course of countless ages, all the
molecules of hydrogen gas would escape from the earth, and in this
manner we may explain the fact that there is no free hydrogen present in
the earth's atmosphere.[10]

The velocities which can be attained by the molecules of gases other
than hydrogen are far too small to permit of their escape from the
attraction of the earth. We therefore find oxygen, nitrogen, water
vapour, and carbon dioxide remaining as permanent components of our air.
On the other hand, the enormous mass of the sun makes the "critical
velocity" at the surface of that body to be so great (391 miles per
second) that not even the molecules of hydrogen can possibly emulate it.
Consequently, as we have seen, hydrogen is a most important component of
the sun's atmospheric envelope.

If we now apply this reasoning to the moon, the critical velocity is
found by calculation to be only a mile and a half per second. This seems
to be well within the maximum velocities attainable by the mo